can no longer use my preferred local model

huggingface / speech-to-speech

Speech To Speech: an effort for an open-sourced and modular GPT4-o

Apache License 2.0

3.54k stars 369 forks source link

can no longer use my preferred local model #50

Open scalar27 opened 2 months ago

scalar27 commented 2 months ago

When I use the --lm_model_name flag and the path to one of my local GGUF model files, I get this error: OSError: Incorrect path_or_model_id: [path to file] . Please provide either the path to a local folder or the repo_id of a model on the Hub.

Last week, this was working so I'm wondering what changed in the code. Thanks.

andimarafioti commented 2 months ago

Which lm are you using? Which lm_model_name? Most likely I broke something 😅 Now that the code structure is more clear and the minimal functionalities are all there, I will start working on adding tests to avoid these issues

scalar27 commented 2 months ago

I've got a folder of about 10 different LLM that I use with llama.cpp. Llama 3, Gemma 2, Hermes, etc.

andimarafioti commented 2 months ago

Can you give me a bit more details? I'm running code with this flag just fine :/